Outcome of Mycobacterium tuberculosis infection in humans is clinically defined as "active" or "latent". These clinical definitions are inadequate to describe the continuum of M. tuberculosis infection. The actual presentation of "active" tuberculosis varies from mild to severe pulmonary disease, including cavitary tuberculosis, and to extrapulmonary or disseminated disease. Several lines of evidence support that latent infection is also a spectrum of infection outcomes, from subclinical disease to "dormant infection" to completely cleared infection. The concept of a latency spectrum has practical implications: we hypothesize that only a small percentage of latently infected persons is most likely to reactive TB, and identifying those persons, who we believe are "higher" on the latency spectrum, allows one to target interventions to those who most will benefit. In this proposal, we will explore the concept of the spectrum of latency and implications for reactivation using a systems biology approach. We propose to integrate data from humans, non-human primates, and computational systems to provide a comprehensive approach to latency and reactivation. We will use immunologic methods and state-of-the-art imaging technology to define the spectrum of latency in humans and non-human primates infected with M. tuberculosis. From non-human primates, we will go one step further and obtain granulomas for detailed study of the spectrum of latency, as well as during reactivation. These granulomas will be used in immunologic, microbiologic and pathologic studies to identify which are most likely to reactivate and the factors involved in maintaining a subclinical infection. In addition to vastly increasing our understanding of "latent" TB and the factors that contribute to reactivation, all of the human and non-human primate data will be incorporated into next generation multi-scale mathematical models of tuberculosis. This will provide the computational platform for sophisticated analysis of factors that contribute to the spectrum of latency and the risk of reactivation. Ultimately these models, informed by data from humans and a very relevant animal model, can be used to test our hypothesis that the position of an individual on the spectrum of latency influences the risk of reactivation. This project brings together an experienced team of immunologists, microbiologists, and computational scientists who have focused on the study of tuberculosis for many years. PUBLIC HEALTH RELEVANCE: Mycobacterium tuberculosis, the causative agent of tuberculosis, can cause clinically apparent disease (TB) or more commonly a clinically silent infection (latent TB) that can reactivate to cause TB as well. It is estimated that 1.7 billion people worldwide have latent TB infection. Here we integrate data from humans and animal models with computational and mathematical models in a comprehensive systems biology approach to a better understanding of latent TB and the factors that lead to reactivation.